Re: flexfax: Whose looking after Hylafax?

[Damian A Ivereigh <damian@cisco.com>]

darren@hylafax.org wrote:
> 
> In message <36F1A81B.6AF1B3C0@cisco.com>, Damian A Ivereigh writes:
> >Would the Hylafax maintainer please stand up!
> 
> That would be, officially, Matthias.
> 
> >Having seen the same question (egcs problem), posted onto the mailing
> >list over and over again, when is this going into the release? What
> >about all the other patches too (e.g. the tag-line patch)?
> 
> I agree with you 100%. And I share your frustration. That's why I registered
> HylaFAX.ORG some time ago now. Fortunately several long-time HylaFAX gurus
> came on-board early, and together we have been able to produce a web site, ftp
> repository and archives which, I feel, are a truly worthwhile service to the
> community.

Absolutely.

> I'm afraid that your post, Damian, has pre-empted the discussions we wanted to
> have with Matthias about "the future". Further exacerbating the situation,
> Matthias is traveling and unable to share with us his thoughts on the matter.
> I think we need to hear from him before anything "official" should/can happen.

That is unfortunate timing. He is back in April, well if it all waits
until then, it can't hurt (we've been waiting enough already :-))

> >David Woolley must spend half his waking life answering questions on
> >the mailing list - he does a sterling effort, but wastes much time
> >with the same questions again and again. No new release to permanently
> >squash the problems.
> 
> We're hoping the archives will alleviate much of this. Kudos to Phil W. for
> some sterling work to bring these to you. Alas, we did not wish to push the
> patches too heavily on the www site because of their "unofficial" nature, and
> also because we have been hoping to obsolete them all with a new release,
> preferably from Matthias, to whom we had intended to submit our work on a new
> patchlevel (Robert is heavily involved in this area).

I will re-iterate all you guys have been doing a great job with the
patches and the web site. I just feel you guys need to get the
confidence to move on to the next step of producing a release.

> >Darren Nickerson did a great job putting together the RPM for Hylafax
> >- but what does he do? Take the original source and apply all the
> >juicy patches to it.
> 
> Thanks, that would be me ;-) Yes, the RPM was the early stages in my
> frustration at the lack of a new patchlevel. At least the linux side of things
> would be up to snuff I thought. Alas, that was not enough . . . the rest of
> the community still lagged behind, and so hylafax.org was born. Progress comes
> slowly my friend, we all have other jobs/responsibilities.

Oh sure, we are all busy. Slow progress isn't the problem per-se, it
is the waste of time and effort which I get fustrated by.

> >Look guys is this stuff Open Source/Free Software or what? What
> >happened to the release early and release often model (aka Bazaar
> >model)?
> 
> Not traditionally the way HylaFAX has been done, alas.

I guess I hadn't fully taken that on board.

> >My understanding is that Sam Leffler, who originally wrote most of the
> >code, no longer had the time or the inclination to continue. It seems
> >he nominated Mathias, be he doesn't have the time to do it either - or
> >at that is how it appears. This is fine, however he must designate a
> >new maintainer.
> 
> Sam no longer works at SGI, and does not intend to participate in HylaFAX
> development in the future. I do not know how Matthias feels about this whole
> thing, he's been a hard man to elicit a response from.

Well that in itself is very telling. A maintainer has to be a leader,
he must be accessible and be prepared to make decisions.

> >I see a lot of good, talented & motivated people running around this
> >list. I think it is time we moved forward. If not, I think it is time
> >we did the unthinkable and forked the code.
> 
> We've been working very hard to prevent a code fork. I understand your
> frustration, honestly, but let's be careful to respect the work of people like
> Matthias, and attempt an inclusive approach, wherever possible. It's a dilemma
> we're attempting to work through. A bit more time will help.

Look I understand why you want to prevent a code fork. Code forks are
always a big hassle generally (understatement), however they should
not be dismissed out of hand. Sometimes they, or the threat of them,
is the only way to move forward. Nothing I say is meant to be a slight
against Matthias' past work. He has done a great job. However he must
avoid undoing that great work, by now getting in the way of its
further progress. I have been in Mathias' shoes, I have seen my users
desert me when I wasn't doing what I should have done. Only after a
friend came to me and suggested, I hand on the baton so to speak, did
I realise what I was doing wrong. However by then it was too late and
the code essentially no longer exists.

> >So guys I think it is time for Hylafax 4.1! Take all those patches -
> >shove them in there and just shove it up on the Hylafax web site. If
> >not, I'm going to put it up on mine. It will create lots of confusion
> >and hassle, but if its the only way to start this project moving
> >again, it'll be worth it.
> 
> This already exists. We'll make it available for testing. Again, you've really
> forced the issue here, perhaps for the greater good, but we're still a bit
> flat-footed on the codebase-maintenance issues.

Well that is what I am hoping. However as a general rule, it is better
to be open about what the code is doing. Even if the code isn't
complete, has bugs, or the docs aren't great. Open it up and solicit
contributions, however you obviously understand this.

> >Its a testament to the quality of the code, that it is still being
> >used despite that lack of maintenance. There is obviosly lots of
> >demand for it. Most code left like this would have died long ago.
> 
> It's a superb product. I'm committed to it with TPC.INT, the other hat I sometimes wear. And I'm highly motivated to see that it continues to grow. For several months now, unknown to most for the reasons hinted at above, we have been patching up the distribution using a CVS repository . . . it's about time what we confessed, came clean, and opened this up to a lot more people.

Yup.

> I'm busy with personal matters over the weekend, but on Sunday or Monday I'll re-enable the anonymous CVS repository (installed and tested some time ago now), and post instructions on how to access it to check out the current kit. Robert may have more to say on this matter, and has advised me that he will prepare a WWW page with all the details once I sort them out.

Hey no panic! We've done without it thus far, a few more days isn't
going to hurt :-)
> 
> I'd like to apologise to you all for the pace at which we've been moving on code-related stuff. I hope you will understand that it is because we are takings thing very seriously indeed, not the converse.

No need to apologise! Just keep everything on track and we'll get
there!

As a general point, I want to re-iterate my thanks to Darren, Phil,
Robert, David Wolley, Mathias and Sam. This is a great piece of code
and it is because I love it, that I am kicking up this fuss. If I
didn't, I would have just sat in the wings and done nothing.

All the best to you all,

Damian

-- 

______________________________________________________________________
* Damian Ivereigh     *      ||        ||      * Cisco Systems, Inc.  
*
* MIS Printer Admin   *    .||||.    .||||.    * Sydney, Australia    
*
* Linux Bigot         * ..:||||||:..:||||||:.. * +61 2 8448 7344      
*
* damian@cisco.com    *   cisco Systems, Inc.  * Fax: +61 2 8448 7228 
*
*______________________________________________________________________*